>>14321553Isn't that quite intuitively explained if one doesn't consider it "changing" the past, but rather "constraining" the past from multiple equally possible pasts? That is, either you know which of the possible pasts happened by measuring a photon at a certain detector, or you don't know which of multiple possible past happened (and then they all must have happened), which means you necessarily also have not have constrained or maybe "focused" the past clearly enough for the photon to have gone only through one or the other slit/path. Otherwise the exact past wouldn't actually be indefinite and you'd know the exact past you're from by checking the other photon's path or the resulting pattern and both can' t be true at the same time. If that counts as intuitive, at least.
Also the significance of consciousness seems overstated? Wouldn't you get the same effect by temporarily saving the results of a few runs in some electronic device or whatever, adding up all the outcomes, then deleting the information of the individual runs and only then comparing the probabilities with what happens when watching the experiment directly? It would be interesting to see how far you could take that indirect measuring before things start "blurring", if ever. That would be analogous to a Wigner's friend scenario, I guess?, but I don't think it's actually been ever done in practice, sadly.
>>14320780tl;dr: The setup and the premise were wrong. Light isn't quantized in terms of energy levels as it travels through space and there was in fact a continuous beam after reducing the energy levels to below what you'd expect from a single photon. Still makes me wonder how a continuous beam of light can interact when it's under the actually quantized interaction threshold, but I suppose it's some tunneling magic or the energy is integrated over time.
See:
https://youtu.be/SDtAh9IwG-I?t=608 
